ABSTRACT In the current era of potent antiretroviral therapy, patients with HIV are well known to develop both infectious (typically Cytomegalovirus Retinitis) as well as non-infectious (retinal microvascular occlusions most often seen as cotton wool spots) forms of retinopathy. There are over one million HIV positive individuals in the US and the disease disproportionately affects minorities. At the present time, non-infectious retinopathy is common and we have shown it is a cause of vision loss even in eyes without overt infectious retinitis. There is structural damage to the inner retina in HIV patients and that this correlates with vision loss and dysfunction. Infectious CMV retinitis is still a cause of severe vision loss due to resistant CMV retinitis in patients who are not responding well or who cannot reliably take HAART therapy. We will use novel methods to analyze retinal structure and function in HIV patients. We will use molecular genetic methods to determine the molecular pathogenesis of retinopathy and plan to use a novel drug delivery system to develop intravitreal therapies for or difficult to treat CMV retinitis. First, we will determine the prevalence and severity of visual dysfunction in HIV patients. We hypothesize that the cumulative effect of these lesions and other areas of retinal vessel damage seen in HIV patients cause widespread retinal damage accounting for the vision loss. We will determine the correlates between retinal structure and function and real world vision performance such as quality of life and driving simulation. Retinal damage will be assessed using novel eye tracking spectral domain OCT technology and microperimetric techniques. Multifocal electrophysiology will determine the level of the retina most involved. For our second aim, autopsy eye mRNA will be analyzed to determine which pathogenic pathways are active in damaged areas. In addition, histologic and morphometric methods will be used to determine the amount of retinal ganglion cell and nerve fiber layer damage and it's correlation with activated pathogenic pathways. Optic nerve degeneration and apoptosis in tissue will also be evaluated. The third aim is to develop an ultra long acting drug delivery system for treatment of resistant CMV retinitis. We have determined that certain derivatives of cidofovir and antiviral acyclic nucleoside phosphonates are highly active against HCMV in vitro even in cases of virus resistant to the usually used anti-CMV compounds. We will be using a novel lipid derivitization method to crystallize these drugs to allow prolonged release and dissolution in the eye. Toxicity and pharmacokinetics will be optimized and be tested in models of retinitis as a precursor to future clinical trials.